Implications of Clouds for eScience and CReSIS CReSIS University of Kansas Lawrence January Geoffrey Fox Director, Digital Science Center, Pervasive Technology Institute Associate Dean for Research and Graduate Studies, School of Informatics and Computing Indiana University Bloomington

Talk Components Important Trends Clouds and Cloud Technologies Applications FutureGrid

Important Trends Data Deluge in all fields of science Multicore implies parallel computing important again – Performance from extra cores – not extra clock speed – GPU enhanced systems can give big power boost Clouds – new commercially supported data center model replacing compute grids (and your general purpose computer center) Light weight clients: Sensors, Smartphones and tablets accessing and supported by backend services in cloud Commercial efforts moving much faster than academia in both innovation and deployment

Gartner 2009 Hype Curve Clouds, Web2.0 Service Oriented Architectures Transformational High Moderate Low Cloud Computing Cloud Web Platforms Media Tablet

Data Centers Clouds & Economies of Scale I Range in size from “edge” facilities to megascale. Economies of scale Approximate costs for a small size center (1K servers) and a larger, 50K server center. Each data center is 11.5 times the size of a football field TechnologyCost in small- sized Data Center Cost in Large Data Center Ratio Network$95 per Mbps/ month $13 per Mbps/ month 7.1 Storage$2.20 per GB/ month $0.40 per GB/ month 5.7 Administration~140 servers/ Administrator >1000 Servers/ Administrator Google warehouses of computers on the banks of the Columbia River, in The Dalles, Oregon Such centers use 20MW-200MW (Future) each with 150 watts per CPU Save money from large size, positioning with cheap power and access with Internet

6 Builds giant data centers with 100,000’s of computers; ~ to a shipping container with Internet access “Microsoft will cram between 150 and 220 shipping containers filled with data center gear into a new 500,000 square foot Chicago facility. This move marks the most significant, public use of the shipping container systems popularized by the likes of Sun Microsystems and Rackable Systems to date.” Data Centers, Clouds & Economies of Scale II

Amazon offers a lot!

X as a Service SaaS: Software as a Service imply software capabilities (programs) have a service (messaging) interface – Applying systematically reduces system complexity to being linear in number of components – Access via messaging rather than by installing in /usr/bin IaaS: Infrastructure as a Service or HaaS: Hardware as a Service – get your computer time with a credit card and with a Web interface PaaS: Platform as a Service is IaaS plus core software capabilities on which you build SaaS Cyberinfrastructure is “Research as a Service” Other Services Clients

Sensors as a Service Cell phones are important sensor Sensors as a Service Sensor Processing as a Service (MapReduce)

C4 = Continuous Collaborative Computational Cloud C4 EMERGING VISION While the internet has changed the way we communicate and get entertainment, we need to empower the next generation of engineers and scientists with technology that enables interdisciplinary collaboration for lifelong learning. Today, the cloud is a set of services that people intently have to access (from laptops, desktops, etc). In 2020 the C4 will be part of our lives, as a larger, pervasive, continuous experience. The measure of success will be how “invisible” it becomes. C4 Education Vision C4 Education will exploit advanced means of communication, for example, “Tabatars” conference tables, with real-time language translation, contextual awareness of speakers, in terms of the area of knowledge and level of expertise of participants to ensure correct semantic translation, and to ensure that people with disabilities can participate. While we are no prophets and we can’t anticipate what exactly will work, we expect to have high bandwidth and ubiquitous connectivity for everyone everywhere, even in rural areas (using power-efficient micro data centers the size of shoe boxes) C4 Society Vision

C 4 Continuous Collaborative Computational Cloud C4C4 I N T E L I G L E N C E Motivating Issues job / education mismatch Higher Ed rigidity Interdisciplinary work Engineering v Science, Little v. Big science Modeling & Simulation C(DE)SE C 4 Intelligent Economy C 4 Intelligent People Stewards of C 4 Intelligent Society NSF Educate “Net Generation” Re-educate pre “Net Generation” in Science and Engineering Exploiting and developing C 4 C 4 Stewards C 4 Curricula, programs C 4 Experiences (delivery mechanism) C 4 REUs, Internships, Fellowships Computational Thinking Internet & Cyberinfrastructure Higher Education 2020

Philosophy of Clouds and Grids Clouds are (by definition) commercially supported approach to large scale computing – So we should expect Clouds to replace Compute Grids – Current Grid technology involves “non-commercial” software solutions which are hard to evolve/sustain – Maybe Clouds ~4% IT expenditure 2008 growing to 14% in 2012 (IDC Estimate) Public Clouds are broadly accessible resources like Amazon and Microsoft Azure – powerful but not easy to customize and perhaps data trust/privacy issues Private Clouds run similar software and mechanisms but on “your own computers” (not clear if still elastic) – Platform features such as Queues, Tables, Databases currently limited Services still are correct architecture with either REST (Web 2.0) or Web Services Clusters are still critical concept for MPI or Cloud software

Grids MPI and Clouds Grids are useful for managing distributed systems – Pioneered service model for Science – Developed importance of Workflow – Performance issues – communication latency – intrinsic to distributed systems – Can never run large differential equation based simulations or datamining Clouds can execute any job class that was good for Grids plus – More attractive due to platform plus elastic on-demand model – MapReduce easier to use than MPI for appropriate parallel jobs – Currently have performance limitations due to poor affinity (locality) for compute-compute (MPI) and Compute-data – These limitations are not “inevitable” and should gradually improve as in July 13 Amazon Cluster announcement – Will probably never be best for most sophisticated parallel differential equation based simulations Classic Supercomputers (MPI Engines) run communication demanding differential equation based simulations – MapReduce and Clouds replaces MPI for other problems – Much more data processed today by MapReduce than MPI (Industry Informational Retrieval ~50 Petabytes per day)

Cloud Computing: Infrastructure and Runtimes Cloud infrastructure: outsourcing of servers, computing, data, file space, utility computing, etc. – Handled through Web services that control virtual machine lifecycles. Cloud runtimes or Platform: tools (for using clouds) to do data- parallel (and other) computations. – Apache Hadoop, Google MapReduce, Microsoft Dryad, Bigtable, Chubby and others – MapReduce designed for information retrieval but is excellent for a wide range of science data analysis applications – Can also do much traditional parallel computing for data-mining if extended to support iterative operations – MapReduce not usually on Virtual Machines

MapReduce Implementations (Hadoop – Java; Dryad – Windows) support: – Splitting of data – Passing the output of map functions to reduce functions – Sorting the inputs to the reduce function based on the intermediate keys – Quality of service Map(Key, Value) Reduce(Key, List ) Data Partitions Reduce Outputs A hash function maps the results of the map tasks to reduce tasks

MapReduce “File/Data Repository” Parallelism Instruments Disks Map 1 Map 2 Map 3 Reduce Communication Map = (data parallel) computation reading and writing data Reduce = Collective/Consolidation phase e.g. forming multiple global sums as in histogram Portals /Users Iterative MapReduce Map Map Reduce Reduce Reduce

All-Pairs Using DryadLINQ Calculate Pairwise Distances (Smith Waterman Gotoh) 125 million distances 4 hours & 46 minutes 125 million distances 4 hours & 46 minutes Calculate pairwise distances for a collection of genes (used for clustering, MDS) Fine grained tasks in MPI Coarse grained tasks in DryadLINQ Performed on 768 cores (Tempest Cluster) Moretti, C., Bui, H., Hollingsworth, K., Rich, B., Flynn, P., & Thain, D. (2009). All-Pairs: An Abstraction for Data Intensive Computing on Campus Grids. IEEE Transactions on Parallel and Distributed Systems, 21,

Hadoop VM Performance Degradation 15.3% Degradation at largest data set size

Cap3 Performance with Different EC2 Instance Types

Cap3 Cost

SWG Cost

Smith Waterman: Daily Effect

Hadoop and Matlab Data Processing: The main application is developed in Matlab, and the data is processed in the following procedure: Simplify the flight lines using Douglas-Peucker algorithm. Export the simplified flight lines as Google KML. Generate radar images. Combine KML and radar images to the self-contained Google KMZ file. Generate the overall simplified flight lines and radar images for the on-going visualization project.

Hadoop and Matlab Sample Output of Flight Line

Hadoop and Matlab 1 Flight Line2 Flight Lines3 Flight Lines PSEPSEPSE Computing nodes % % % % % % % % % % % % % % %0.81 Hadoop Test Results. P=Processing time (s). S=T(3 nodes)/T (x nodes) relative speedup. E=Efficiency per node, S(nodes)*3/nodes

Matlab and Hadoop Speed up versus number of nodes

Visualization of Cloud Data Cross-section View of 2 Surfaces

Volume Data: Extract Cross Sections

US Cyberinfrastructure Context There are a rich set of facilities – Production TeraGrid facilities with distributed and shared memory – Experimental “Track 2D” Awards FutureGrid: Distributed Systems experiments cf. Grid5000 Keeneland: Powerful GPU Cluster Gordon: Large (distributed) Shared memory system with SSD aimed at data analysis/visualization – Open Science Grid aimed at High Throughput computing and strong campus bridging

30 TeraGrid ‘10 August 2-5, 2010, Pittsburgh, PA SDSC TACC UC/ANL NCSA ORNL PU IU PSC NCAR Caltech USC/ISI UNC/RENCI UW Resource Provider (RP) Software Integration Partner Grid Infrastructure Group (UChicago) TeraGrid ~2 Petaflops; over 20 PetaBytes of storage (disk and tape), over 100 scientific data collections NICS LONI Network Hub

FutureGrid key Concepts I FutureGrid is an international testbed modeled on Grid5000 Supporting international Computer Science and Computational Science research in cloud, grid and parallel computing – Industry and Academia – Prototype software development and Education/Training – Mainly computer science, bioinformatics, education The FutureGrid testbed provides to its users: – A flexible development and testing platform for middleware and application users looking at interoperability, functionality and performance, exploring new computing paradigms – Each use of FutureGrid is an experiment that is reproducible – A rich education and teaching platform for advanced cyberinfrastructure classes – Support for users experimentation

FutureGrid key Concepts II Rather than loading images onto VM’s, FutureGrid supports Cloud, Grid and Parallel computing environments by dynamically provisioning software as needed onto “bare-metal” using Moab/xCAT –Image library for all the different environments you might like to explore ….. Growth comes from users depositing novel images in library FutureGrid has ~4000 (will grow to ~5000) distributed cores with a dedicated network and a Spirent XGEM network fault and delay generator Apply now to use FutureGrid on web site Image1 Image2 ImageN … LoadChooseRun

FutureGrid and clouds for CReSIS? Clouds could be used by CReSIS in – Research – Education FutureGrid can be vehicle for – Supporting CS Research – Experimenting with cloud approaches for CReSIS data analysis We could set up a customized ongoing support activity on FutureGrid for CReSIS We could offer a hands-on tutorial or summer school – See Jerome Mitchell proposal ADMI could use FutureGrid well

FutureGrid Partners Indiana University (Architecture, core software, Support) – Collaboration between research and infrastructure groups Purdue University (HTC Hardware) San Diego Supercomputer Center at University of California San Diego (INCA, Monitoring) University of Chicago/Argonne National Labs (Nimbus) University of Florida (ViNE, Education and Outreach) University of Southern California Information Sciences (Pegasus to manage experiments) University of Tennessee Knoxville (Benchmarking) University of Texas at Austin/Texas Advanced Computing Center (Portal) University of Virginia (OGF, Advisory Board and allocation) Center for Information Services and GWT-TUD from Technische Universtität Dresden. (VAMPIR) Red institutions have FutureGrid hardware

Compute Hardware System type# CPUs# CoresTFLOPS Total RAM (GB) Secondary Storage (TB) Site Status IBM iDataPlex *IU Operational Dell PowerEdge TACC Operational IBM iDataPlex UC Operational IBM iDataPlex SDSC Operational Cray XT5m *IU Operational IBM iDataPlex On OrderUF Operational Large disk/memory system TBD on nodesIU New System TBD High Throughput Cluster PU Not yet integrated Total

FutureGrid: a Grid/Cloud/HPC Testbed Private Public FG Network NID : Network Impairment Device

37 Typical Performance Study Linux, Linux on VM, Windows, Azure, Amazon Bioinformatics

Some Current FutureGrid projects

OGF’10 Demo SDSC UF UC Lille Rennes Sophia ViNe provided the necessary inter-cloud connectivity to deploy CloudBLAST across 5 Nimbus sites, with a mix of public and private subnets. Grid’5000 firewall

University of Arkansas Indiana University University of California at Los Angeles Penn State Iowa Univ.Illinois at Chicago University of Minnesota Michigan State Notre Dame University of Texas at El Paso IBM Almaden Research Center Washington University San Diego Supercomputer Center University of Florida Johns Hopkins July 26-30, 2010 NCSA Summer School Workshop Students learning about Twister & Hadoop MapReduce technologies, supported by FutureGrid.

User Support Being upgraded now as we get into major use Regular support: there is a group forming FET or “FutureGrid Expert Team” – initially 13 PhD students and researchers from Indiana University – User requests project at account-project-registrationhttp:// account-project-registration – Each user assigned a member of FET when project approved – Users given accounts when project approved – FET member and user interact to get going on FutureGrid – Could have identified CReSIS or ADMI support people Advanced User Support: limited special support available on request – Cummins engine simulation supported in this way